[{"data":1,"prerenderedAt":-1},["ShallowReactive",2],{"project-74742":3},{"id":4,"name":5,"fullName":6,"owner":7,"repo":5,"description":8,"homepage":9,"htmlUrl":10,"language":11,"languages":10,"totalLinesOfCode":10,"stars":12,"forks":13,"watchers":14,"openIssues":15,"contributorsCount":16,"subscribersCount":16,"size":16,"stars1d":17,"stars7d":18,"stars30d":19,"stars90d":16,"forks30d":16,"starsTrendScore":20,"compositeScore":21,"rankGlobal":10,"rankLanguage":10,"license":22,"archived":23,"fork":23,"defaultBranch":24,"hasWiki":25,"hasPages":25,"topics":26,"createdAt":10,"pushedAt":10,"updatedAt":37,"readmeContent":38,"aiSummary":39,"trendingCount":16,"starSnapshotCount":16,"syncStatus":40,"lastSyncTime":41,"discoverSource":42},74742,"typewhisper-mac","TypeWhisper\u002Ftypewhisper-mac","TypeWhisper","Local speech-to-text for macOS  on-device AI, fully private, optional cloud","",null,"Swift",1342,92,5,26,0,21,56,255,63,18.91,"GNU General Public License v3.0",false,"main",true,[27,28,29,30,31,32,33,34,35,36],"apple-silicon","dictation","macos","on-device","privacy","speech-to-text","swiftui","transcription","whisper","whisperkit","2026-06-12 02:03:27","# TypeWhisper for Mac\n\n[![License: GPL v3](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-GPLv3-blue.svg)](https:\u002F\u002Fwww.gnu.org\u002Flicenses\u002Fgpl-3.0)\n[![macOS](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FmacOS-14.0%2B-black.svg)](https:\u002F\u002Fwww.apple.com\u002Fmacos\u002F)\n[![Swift](https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSwift-6-orange.svg)](https:\u002F\u002Fswift.org)\n\nSpeech-to-text and AI text processing for macOS. Transcribe audio using on-device AI models or cloud APIs (Groq, OpenAI, xAI\u002FGrok), then transform the result with reusable workflows. Your voice data stays on your Mac with local models - or use cloud APIs for faster processing.\n\nTypeWhisper `1.4` is the current release-candidate line for macOS. It includes system-wide dictation, file transcription, unified workflows, history, dictionary, snippets, and bundled integrations. Advanced surfaces like the HTTP API, CLI, widgets, watch folders, and the plugin SDK remain supported for power users and automation.\n\nSee the [release readiness guide](docs\u002Frelease-readiness.md), [support matrix](docs\u002Fsupport-matrix.md), and [release checklist](docs\u002Frelease-checklist.md) for the current release definition and ship gates.\n\n\u003Cp align=\"center\">\n  \u003Cvideo src=\"https:\u002F\u002Fgithub.com\u002Fuser-attachments\u002Fassets\u002F22fe922d-4a4c-47d1-805e-684a148ebd03\" autoplay loop muted playsinline width=\"270\">\u003C\u002Fvideo>\n\u003C\u002Fp>\n\n## Screenshots\n\n\u003Cp align=\"center\">\n  \u003Ca href=\".github\u002Fscreenshots\u002Fhome.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fhome.png\" width=\"270\" alt=\"Home Dashboard\">\u003C\u002Fa>\n  \u003Ca href=\".github\u002Fscreenshots\u002Frecording.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Frecording.png\" width=\"270\" alt=\"Recording & Hotkeys\">\u003C\u002Fa>\n  \u003Ca href=\".github\u002Fscreenshots\u002Fworkflows.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fworkflows.png\" width=\"270\" alt=\"Workflows\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\".github\u002Fscreenshots\u002Fhistory.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fhistory.png\" width=\"270\" alt=\"Transcription History\">\u003C\u002Fa>\n  \u003Ca href=\".github\u002Fscreenshots\u002Fdictionary.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fdictionary.png\" width=\"270\" alt=\"Dictionary\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\".github\u002Fscreenshots\u002Fgeneral.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fgeneral.png\" width=\"270\" alt=\"General Settings\">\u003C\u002Fa>\n  \u003Ca href=\".github\u002Fscreenshots\u002Fplugins.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fplugins.png\" width=\"270\" alt=\"Integrations\">\u003C\u002Fa>\n  \u003Ca href=\".github\u002Fscreenshots\u002Ffile-transcription.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Ffile-transcription.png\" width=\"270\" alt=\"File Transcription\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n\u003Cp align=\"center\">\n  \u003Ca href=\".github\u002Fscreenshots\u002Fsnippets.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fsnippets.png\" width=\"270\" alt=\"Snippets\">\u003C\u002Fa>\n  \u003Ca href=\".github\u002Fscreenshots\u002Fadvanced.png\">\u003Cimg src=\".github\u002Fscreenshots\u002Fadvanced.png\" width=\"270\" alt=\"Advanced Settings\">\u003C\u002Fa>\n\u003C\u002Fp>\n\n## What's New in 1.3\n\n- **Unified Workflows** - Prompt actions and matching rules now live in one dedicated workflow surface with a native editor and guided creation flow\n- **Always fallback trigger** - Create a global workflow that runs when no more specific app or website workflow matches\n- **Manual workflow palette** - Keep workflows out of automatic dictation matching and run them from one global Workflow Palette shortcut\n- **Safer prompt boundaries** - Workflow prompts treat dictated text as source content to transform, not instructions to execute\n- **Focus-safe local processing** - On-device workflows keep focus in the original target app instead of foregrounding TypeWhisper unexpectedly\n- **Snippets and dictionary polish** - Snippets are first-class settings, dictionary term packs are easier to review, and corrections stay engine-aware\n- **Integration refresh** - Bundled transcription, LLM, and action plugins are easier to inspect and activate\n\n## Features\n\n### Transcription\n\n- **Ten engines** - WhisperKit (99+ languages, streaming, translation), Parakeet TDT v3 (25 European languages, extremely fast), Apple SpeechAnalyzer (macOS 26+, no model download needed), Granite Speech (MLX-based), Qwen3 ASR (MLX-based), Voxtral (local Voxtral Mini 4B, MLX-based), Groq Whisper, OpenAI Whisper, xAI\u002FGrok STT, and OpenAI Compatible (any OpenAI-compatible API)\n- **On-device or cloud** - All processing happens locally on your Mac, or use Groq\u002FOpenAI\u002FxAI APIs for faster processing\n- **Streaming preview** - See partial transcription in real-time while speaking (WhisperKit)\n- **Short-clip handling** - Better retention of brief utterances and fewer false no-speech discards\n- **File transcription** - Batch-process multiple audio\u002Fvideo files with drag & drop\n- **Subtitle export** - Export transcriptions as SRT or WebVTT with timestamps\n\n### Dictation\n\n- **System-wide** - Push-to-talk, toggle, or hybrid mode via global hotkey, auto-pastes into any app\n- **Modifier-key hotkeys** - Use a single modifier key (Command, Shift, Option, Control) as your hotkey\n- **Indicator styles** - Choose Notch, Overlay, or Minimal, with optional live transcript preview where supported\n- **Sound feedback** - Audio cues for recording start, transcription success, and errors\n- **Microphone selection** - Choose a specific input device with live preview and improved recovery after route changes\n\n### AI Processing\n\n- **Workflows** - Build reusable transformations for translation, rewriting, extraction, formatting, and app-specific automation. Workflows can run automatically by app, website, or app + website combinations, from a dedicated hotkey, as a global fallback, or manually from the Workflow Palette. Hotkey workflows can either start dictation or process the current selection\u002Fclipboard directly.\n- **LLM providers** - Apple Intelligence (macOS 26+), Groq, OpenAI \u002F ChatGPT, xAI\u002FGrok, Gemini, and OpenAI Compatible with per-prompt provider and model override\n- **Speech providers** - System voices, xAI\u002FGrok TTS, and experimental local Supertonic TTS can provide spoken feedback and readback\n- **Local prompt processing** - Gemma 4 via MLX runs on-device on Apple Silicon, with the current verified release path limited to the E2B\u002FE4B 4-bit models\n- **Translation** - Translate transcriptions on-device using Apple Translate\n\n### Personalization\n\n- **Workflow triggers** - Per-app, per-website, combined app + website, hotkey, global fallback, and manual palette-only triggers for language, task, engine, prompt, and auto-submit behavior. Website matching supports subdomains\n- **Dictionary** - Terms improve cloud recognition accuracy. Corrections fix common transcription mistakes automatically. Auto-learns from manual corrections. Includes importable term packs\n- **Localized term packs** - Built-in term pack names and descriptions are localized in English and German\n- **Snippets** - Text shortcuts with trigger\u002Freplacement. Supports placeholders like `{{DATE}}`, `{{TIME}}`, and `{{CLIPBOARD}}`\n- **History** - Searchable transcription history with inline editing, correction detection, app context tracking, timeline grouping, filters, bulk delete, multi-select export, auto-retention, and a standalone window accessible from the tray menu\n\n### Integration & Extensibility\n\n- **Plugin system** - Extend TypeWhisper with custom LLM providers, transcription engines, TTS providers, post-processors, and action plugins. Granite, Groq, OpenAI \u002F ChatGPT, OpenAI Compatible, xAI\u002FGrok, Gemini, Linear, Qwen3, Supertonic, Voxtral, and Webhook ship as bundled plugins, alongside the local engine plugins. Linear plugin enables voice-to-issue creation. See [TypeWhisperPluginSDK\u002FPlugins\u002FREADME.md](TypeWhisperPluginSDK\u002FPlugins\u002FREADME.md)\n- **Local model download controls** - Bundled Qwen3, Granite, Voxtral, and Supertonic plugins support an optional HuggingFace token for higher rate limits and clearer download errors. Supertonic requires explicit OpenRAIL-M model-license acceptance before model assets download.\n- **HTTP API** - Local REST API for integration with external tools and scripts\n- **CLI tool** - Shell-friendly transcription via the command line\n- **Discord claim service** - Optional external service for Polar supporter and GitHub Sponsors Discord role claims\n\n### General\n\n- **Home dashboard** - Usage statistics, activity chart, and onboarding tutorial\n- **Auto-update** - Built-in updates via Sparkle with stable, release-candidate, and daily channels\n- **Universal binary** - Runs natively on Apple Silicon and Intel Macs\n- **Widgets** - Desktop widgets for usage stats, last transcription, activity chart, and transcription history\n- **Multilingual UI** - English and German\n- **Launch at Login** - Start automatically with macOS\n\n## Install\n\n### Homebrew\n\n```bash\nbrew install --cask typewhisper\u002Ftap\u002Ftypewhisper\n```\n\n### Direct Download\n\nDownload the latest DMG from [GitHub Releases](https:\u002F\u002Fgithub.com\u002FTypeWhisper\u002Ftypewhisper-mac\u002Freleases\u002Flatest).\n\nStable direct-download releases use the default Sparkle channel. Release candidates such as `1.4.0-rc*` and daily builds are published as GitHub prereleases, update the shared Sparkle appcast on their own channels, and are excluded from Homebrew.\nInstalled builds can switch channels in `Settings -> About` via the `Update Channel` picker.\n\n## Quick Start\n\n1. Install TypeWhisper from Homebrew or the latest DMG.\n2. Open Settings and grant Microphone plus Accessibility access.\n3. Pick an engine and, if needed, download a local model.\n4. Trigger the global hotkey and complete your first dictation.\n\n## Manual Uninstall (macOS)\n\nThese steps are for official TypeWhisper release builds on macOS. They remove the app itself, its local state, widget data, and stored secrets so you can reinstall from a clean slate.\n\nIf you installed via Homebrew, you can optionally start with:\n\n```bash\nbrew uninstall --cask typewhisper\n```\n\nThat removes the app bundle, but it does not reliably remove all files in `~\u002FLibrary` or TypeWhisper entries in Keychain.\n\nIf `~\u002FLibrary` is hidden in Finder, use `Go -> Go to Folder...` and paste the paths below.\n\n1. Quit TypeWhisper if it is running.\n2. Delete the app bundle:\n   ```bash\n   rm -rf \u002FApplications\u002FTypeWhisper.app\n   ```\n3. Delete app data and plugins:\n   ```bash\n   rm -rf ~\u002FLibrary\u002FApplication\\ Support\u002FTypeWhisper\n   ```\n4. Delete preferences:\n   ```bash\n   rm -f ~\u002FLibrary\u002FPreferences\u002Fcom.typewhisper.mac.plist\n   ```\n5. Delete widget and app group data used by official releases:\n   ```bash\n   rm -rf ~\u002FLibrary\u002FGroup\\ Containers\u002F2D8ALY3LCL.com.typewhisper.mac\n   ```\n6. Remove TypeWhisper secrets from Keychain:\n   - In Keychain Access, search for `com.typewhisper.mac.apikey` and delete matching items.\n   - This includes API and plugin secrets stored under the `com.typewhisper.mac.apikey.*` service prefix.\n   - Also remove the license items stored under service `com.typewhisper.mac.apikey.license`, especially the `polar-license` and `polar-supporter` accounts.\n7. If you installed the CLI tool from Settings > Advanced, remove it too:\n   ```bash\n   rm -f \u002Fusr\u002Flocal\u002Fbin\u002Ftypewhisper\n   ```\n8. Optional: if you want to remove exported user files as well, delete:\n   ```bash\n   rm -rf ~\u002FDocuments\u002FTypeWhisper\\ Recordings\n   ```\n9. Restart your Mac, then install the latest build again.\n\nIf a fresh install still crashes immediately after these steps, please open an issue and include your macOS version, how you installed TypeWhisper, and whether the crash happens on first launch or after granting permissions.\n\n## System Requirements\n\n- macOS 14.0 (Sonoma) or later\n- Apple Silicon (M1 or later) recommended\n- 8 GB RAM minimum, 16 GB+ recommended for larger models\n- Some features (Apple Translate, improved Settings UI) require macOS 15+. Apple Intelligence and SpeechAnalyzer require macOS 26+.\n\n## Gemma 4 Support\n\nTypeWhisper includes a bundled local Gemma 4 plugin powered by MLX for on-device prompt processing on Apple Silicon. In the current verified release path, Gemma 4 support is limited to the dense `E2B 4-bit` and `E4B 4-bit` variants; larger or unverified variants stay visible in the UI but remain disabled until they are validated end to end.\n\n## Model Recommendations\n\n| RAM | Recommended Models |\n|-----|-------------------|\n| \u003C 8 GB | Whisper Tiny, Whisper Base |\n| 8-16 GB | Whisper Small, Whisper Large v3 Turbo, Parakeet TDT v3, Voxtral Mini 4B |\n| > 16 GB | Whisper Large v3 |\n\n## Build\n\n1. Clone the repository:\n   ```bash\n   git clone https:\u002F\u002Fgithub.com\u002FTypeWhisper\u002Ftypewhisper-mac.git\n   cd typewhisper-mac\n   ```\n\n2. Open in Xcode 16+:\n   ```bash\n   open TypeWhisper.xcodeproj\n   ```\n\n3. Select the TypeWhisper scheme and build (Cmd+B). Swift Package dependencies (WhisperKit, FluidAudio, Sparkle, TypeWhisperPluginSDK) resolve automatically.\n\n4. Run the app. It appears as a menu bar icon - open Settings to download a model.\n\n5. Run the automated checks before shipping changes:\n   ```bash\n   xcodebuild test -project TypeWhisper.xcodeproj -scheme TypeWhisper -destination 'platform=macOS,arch=arm64' -parallel-testing-enabled NO CODE_SIGN_IDENTITY='-' CODE_SIGNING_REQUIRED=NO CODE_SIGNING_ALLOWED=NO\n   swift test --package-path TypeWhisperPluginSDK\n   ```\n\n## HTTP API\n\nThe HTTP API is an advanced local automation surface. It binds to `127.0.0.1` only, is disabled by default, and is intended for local tools and scripts.\n\nEnable the API server in Settings > Advanced (default port: `8978`).\n\n### Check Status\n\n```bash\ncurl http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fstatus\n```\n\n```json\n{\n  \"status\": \"ready\",\n  \"engine\": \"whisper\",\n  \"model\": \"openai_whisper-large-v3_turbo\",\n  \"supports_streaming\": true,\n  \"supports_translation\": true\n}\n```\n\n### Transcribe Audio\n\n```bash\ncurl -X POST http:\u002F\u002Flocalhost:8978\u002Fv1\u002Ftranscribe \\\n  -F \"file=@recording.wav\" \\\n  -F \"language=en\"\n\ncurl -X POST http:\u002F\u002Flocalhost:8978\u002Fv1\u002Ftranscribe \\\n  -F \"file=@recording.wav\" \\\n  -F \"language_hint=de\" \\\n  -F \"language_hint=en\"\n```\n\n```json\n{\n  \"text\": \"Hello, world!\",\n  \"language\": \"en\",\n  \"duration\": 2.5,\n  \"processing_time\": 0.8,\n  \"engine\": \"whisper\",\n  \"model\": \"openai_whisper-large-v3_turbo\"\n}\n```\n\nOptional parameters:\n- `language` - ISO 639-1 code (e.g., `en`, `de`). Omit for full auto-detection.\n- `language_hint` - Repeatable language hint for restricted auto-detection. Do not combine with `language`.\n- `task` - `transcribe` (default) or `translate` (translates to English, WhisperKit only).\n- `target_language` - ISO 639-1 code for translation target language (e.g., `es`, `fr`). Uses Apple Translate.\n\nUploads to `\u002Fv1\u002Ftranscribe` are limited to 256 MiB, including stdin uploads from the CLI. Requests above that size return `413 Payload Too Large`. Local CLI file paths use a direct handoff to the running TypeWhisper app instead of uploading the file bytes.\n\n### List Models\n\n```bash\ncurl http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fmodels\n```\n\n```json\n{\n  \"models\": [\n    {\n      \"id\": \"openai_whisper-large-v3_turbo\",\n      \"engine\": \"whisper\",\n      \"ready\": true\n    }\n  ]\n}\n```\n\n### History\n\n```bash\n# Search history\ncurl \"http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fhistory?q=meeting&limit=10&offset=0\"\n\n# Delete entry\ncurl -X DELETE \"http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fhistory?id=\u003Cuuid>\"\n```\n\n### Workflows\n\n```bash\n# List all workflow-backed rules\ncurl http:\u002F\u002Flocalhost:8978\u002Fv1\u002Frules\n\n# Toggle a workflow-backed rule on\u002Foff\ncurl -X PUT \"http:\u002F\u002Flocalhost:8978\u002Fv1\u002Frules\u002Ftoggle?id=\u003Cuuid>\"\n```\n\n### Dictation Control\n\n```bash\n# Start dictation (returns session id)\ncurl -X POST http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fdictation\u002Fstart\n\n# Stop dictation (returns same session id)\ncurl -X POST http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fdictation\u002Fstop\n\n# Check whether dictation is currently recording\ncurl http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fdictation\u002Fstatus\n\n# Fetch status\u002Fresult for a specific dictation session\ncurl \"http:\u002F\u002Flocalhost:8978\u002Fv1\u002Fdictation\u002Ftranscription?id=\u003Cuuid>\"\n```\n\n## CLI Tool\n\nTypeWhisper includes a command-line tool for shell-friendly transcription. It is part of the advanced automation surface and connects to the running local API server.\n\n### Installation\n\nInstall via Settings > Advanced > CLI Tool > Install. This places the `typewhisper` binary in `\u002Fusr\u002Flocal\u002Fbin`.\n\n### Commands\n\n```bash\ntypewhisper status              # Show server status\ntypewhisper models              # List available models\ntypewhisper transcribe file.wav # Transcribe an audio file\n```\n\n### Options\n\n| Option | Description |\n|--------|-------------|\n| `--port \u003CN>` | Server port (default: auto-detect) |\n| `--json` | Output as JSON |\n| `--language \u003Ccode>` | Source language (e.g. `en`, `de`) |\n| `--language-hint \u003Ccode>` | Repeatable language hint for restricted auto-detection |\n| `--task \u003Ctask>` | `transcribe` (default) or `translate` |\n| `--translate-to \u003Ccode>` | Target language for translation |\n\n### Examples\n\n```bash\n# Transcribe with language and JSON output\ntypewhisper transcribe recording.wav --language de --json\n\n# Restrict auto-detection to a shortlist\ntypewhisper transcribe recording.wav --language-hint de --language-hint en\n\n# Pipe audio from stdin\ncat audio.wav | typewhisper transcribe -\n\n# Use in a script\ntypewhisper transcribe meeting.m4a --json | jq -r '.text'\n```\n\nThe CLI requires the API server to be running (Settings > Advanced) and follows the documented command and flag surface for the current stable release.\n\nLocal file paths are handed to the running TypeWhisper app directly, so large files do not need to fit inside an HTTP upload body. Stdin usage (`typewhisper transcribe -`) still uses the regular `\u002Fv1\u002Ftranscribe` upload endpoint and is limited to 256 MiB.\n\n## Workflows\n\nWorkflows let you configure transcription, transformation, and automation behavior per application, website, combined app + website context, hotkey, global fallback, or manual palette-only workflow. For example:\n\n- **Mail** - German language, Whisper Large v3\n- **Slack** - English language, Parakeet TDT v3\n- **Terminal** - English language, auto-submit enabled\n- **github.com** - English cleanup workflow that matches in any browser\n- **docs.google.com** - German dictation workflow that translates to English\n\nCreate workflows in Settings > Workflows. Choose a template, then use Automatic to enable app, website, hotkey, or any combination of those trigger components. Always stays the global fallback, and Manual keeps the workflow palette-only. Hotkey workflows choose whether the shortcut starts dictation or processes the current selection\u002Fclipboard through the same insertion path as the Workflow Palette. Spoken language can be left on full auto-detect, fixed to one exact language, or restricted to a shortlist of likely languages for better detection accuracy. Website patterns support subdomain matching - e.g. `google.com` also matches `docs.google.com`.\n\nWhen you start dictating, TypeWhisper matches the active app and browser URL against enabled workflows with the following priority:\n1. **App + URL match** - highest specificity (e.g. Chrome + github.com)\n2. **URL-only match** - cross-browser workflows (e.g. github.com in any browser)\n3. **App-only match** - generic app workflows (e.g. all of Chrome)\n4. **Always fallback** - global workflow when no more specific workflow matches\n\nHotkeys are direct workflow shortcuts, not context conditions in the app\u002FURL matching order. Manual workflows are excluded from automatic dictation matching. They appear only in the Workflow Palette and use the existing Workflow Palette hotkey.\n\nThe active workflow name is shown as a badge in the indicator, together with a short explanation of why it matched.\n\nMultiple engines can be loaded simultaneously for instant switching between workflows. Note that loading multiple local models increases memory usage. Cloud engines (Groq, OpenAI, xAI\u002FGrok) have negligible memory overhead.\n\n## Plugins\n\nTypeWhisper supports plugins for adding custom LLM providers, transcription engines, TTS providers, post-processors, and action plugins. Plugins are macOS `.bundle` files placed in `~\u002FLibrary\u002FApplication Support\u002FTypeWhisper\u002FPlugins\u002F`.\n\nBundled engines and integrations (WhisperKit, Parakeet, SpeechAnalyzer, Granite, Qwen3, Voxtral, Supertonic, Groq, OpenAI, xAI\u002FGrok, OpenAI Compatible, Gemini, Linear, Webhook, and more) are implemented as plugins and serve as reference implementations.\n\nSee [TypeWhisperPluginSDK\u002FPlugins\u002FREADME.md](TypeWhisperPluginSDK\u002FPlugins\u002FREADME.md) for the full plugin development guide, including the event bus, host services API, and manifest format.\n\n## Architecture\n\n```\nTypeWhisper\u002F\n├── typewhisper-cli\u002F           # Command-line tool (status, models, transcribe)\n├── PluginRegistry\u002F            # Source registry entries for community plugin feeds\n├── Plugins\u002F                # Redirect docs and legacy entrypoint for moved first-party plugin sources\n├── TypeWhisperPluginSDK\u002F   # Plugin SDK (Swift package)\n│   ├── Plugins\u002F            # First-party plugin sources and manifests\n├── TypeWhisperWidgetExtension\u002F # WidgetKit widgets (stats, activity, history)\n├── TypeWhisperWidgetShared\u002F    # Shared widget data models\n├── App\u002F                    # App entry point, dependency injection\n├── Models\u002F                 # Data models (TranscriptionResult, Profile, PromptAction, etc.)\n├── Services\u002F\n│   ├── Cloud\u002F              # KeychainService, WavEncoder (shared cloud utilities)\n│   ├── LLM\u002F               # Apple Intelligence provider (cloud LLM providers are plugins)\n│   ├── HTTPServer\u002F         # Local REST API (HTTPServer, APIRouter, APIHandlers)\n│   ├── ModelManagerService # Transcription dispatch (delegates to plugins)\n│   ├── AudioRecordingService\n│   ├── AudioFileService    # Audio\u002Fvideo - 16kHz PCM conversion\n│   ├── HotkeyService\n│   ├── TextInsertionService\n│   ├── WorkflowService     # Workflow matching and persistence\n│   ├── HistoryService      # Transcription history persistence (SwiftData)\n│   ├── DictionaryService   # Custom term corrections\n│   ├── SnippetService      # Text snippets with placeholders\n│   ├── PromptActionService # Prompt action persistence (SwiftData)\n│   ├── PromptProcessingService # LLM orchestration for prompt execution\n│   ├── PluginManager       # Plugin discovery, loading, and lifecycle\n│   ├── PluginRegistryService # Plugin marketplace (download, install, update)\n│   ├── PostProcessingPipeline # Priority-based text processing chain\n│   ├── EventBus            # Typed publish\u002Fsubscribe event system\n│   ├── TranslationService  # On-device translation via Apple Translate\n│   ├── SubtitleExporter    # SRT\u002FVTT export\n│   └── SoundService        # Audio feedback for recording events\n├── ViewModels\u002F             # MVVM view models with Combine\n├── Views\u002F                  # SwiftUI views\n└── Resources\u002F              # Info.plist, entitlements, localization, sounds\n```\n\n**Patterns:** MVVM with `ServiceContainer` singleton for dependency injection. ViewModels use a static `_shared` pattern. Localization via `String(localized:)` with `Localizable.xcstrings`.\n\n## License\n\nGPLv3 - see [LICENSE](LICENSE) for details. Commercial licensing available - see [LICENSE-COMMERCIAL.md](LICENSE-COMMERCIAL.md).\n","TypeWhisper for Mac 是一款专为 macOS 设计的语音转文字工具，支持本地设备上的 AI 处理和可选的云端服务。其核心功能包括系统级听写、文件转录、统一的工作流管理、历史记录、字典和代码片段等。通过使用苹果芯片上的本地 AI 模型，用户的数据可以完全保留在设备上，确保隐私安全；同时，也可以选择使用 Groq、OpenAI 或 xAI\u002FGrok 等云 API 来加速处理过程。该应用适合需要高效且私密的语音输入解决方案的用户，如作家、程序员或任何频繁使用语音输入的人士。此外，它还提供了丰富的高级功能，如 HTTP API、命令行界面、小部件、观察文件夹以及插件 SDK，满足高级用户的需求。",2,"2026-06-11 03:50:36","high_star"]